factor-research

To calculate IC and IR for cross-sectional financial factors, you provide a factor values CSV and a forward-return CSV, then run the factor_analysis tool to generate IC metrics and equity curves. This computes IC/IR across instruments and time to quantify predictive power.

Running a quantile backtest evaluates factor monotonicity and long-short effectiveness by grouping instruments into quantiles based on factor values. The factor_analysis tool processes your input CSVs and outputs group_equity.csv to reveal performance across quantiles.

You need a cross-sectional factor values CSV and a forward-return CSV as inputs for factor screening and backtesting. The process outputs ic_series.csv, ic_summary.json, and group_equity.csv to support your analysis decisions.

Yes, you can test financial factors across multi-instrument universes and various holding periods. The factor_analysis tool applies IC/IR testing and quantile backtests across these dimensions to screen factors and generate robust multi-factor signals.

To combine multiple financial factors into a robust signal, the factor_analysis tool screens individual factors using IC/IR analysis and quantile backtesting results. This validates predictive power and avoids look-ahead bias when building multi-factor strategies.

Factor backtests suffer from look-ahead bias when future data leaks into current period calculations. Using structured IC/IR analysis with properly aligned forward-return CSVs ensures factor values are evaluated strictly against subsequent returns to prevent this.